Skip to main content
Back to challenges
Challenge

Multi-Agent Crude Allocation & Tender Bidding System using Microsoft AutoGen

Managing crude oil allocation between state marketing directives, local refiners, and international trading houses requires dynamic multi-turn negotiations. Build a multi-agent bidding and quota allocation system using Microsoft AutoGen. The system consists of three conversing agents: SOMO Quota Manager, International Trader Representative, and Local Refinery Liaison. The agents converse autonomously to clear spot cargo pricing differentials and confirm allocation schedules based on the approved 3-month mechanism.

Business OperationsHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Develop an autonomous multi-agent crude allocation and tender negotiation graph using Microsoft AutoGen.

Delivery guide

How work is evaluated

Evaluation

Evaluates outcome consensus and constraint compliance in multi-agent tender negotiations.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

agreement_validity

Ensure agreement is reached and highest valid premium wins.

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

negotiation_efficiency

Number of conversation turns required to reach consensus. • target: 4 • range: 1-10

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Orchestrate multi-agent group chats using Microsoft AutoGen framework.

  • Define distinct system messages and negotiation strategies per agent persona.

  • Implement conversation termination conditions based on agreement consensus.

Resources and assets

Reference links and supporting material

Dataset notes

Simulated oil export tender parameters and trader willingness-to-pay profiles.

Challenge setup

Evaluates outcome consensus and constraint compliance in multi-agent tender negotiations.

Challenge input

JSON containing cargo_volume, base_osp_differential, bidder_profiles

AutoGen

Microsoft's multi-agent framework for collaborative task execution.

Evaluated output

JSON with winning_bidder, final_price_differential, agreement_reached

Evaluation checks
  • Ensure agreement is reached and highest valid premium wins.
  • Number of conversation turns required to reach consensus. • target: 4 • range: 1-10
Configured evaluation evidence
  • Benchmark: AutoGen-Market-Bench
  • Negotiation Efficiency target: 4
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • AutoGen · Required
  • Microsoft AutoGen · Optional
  • AutoGluon · Optional

Evaluation contract

  • agreement_validity · Weight 1
  • negotiation_efficiency · Weight 1

Metrics

  • negotiation_efficiency · Target 4

Benchmarks

  • AutoGen-Market-Bench
  • energy

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing cargo_volume, base_osp_differential, bidder_profiles

    Output: JSON with winning_bidder, final_price_differential, agreement_reached

    Execution harness: Python

Evaluation tools

  • AutoGen · Multi-Agent Conversation Engine

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Multi-Agent Crude Allocation & Tender Bidding System using Microsoft AutoGen